OCT close to detecting progression of visual field damage

January 15, 2008

Analysis programs for longitudinal assessment of glaucoma are advancing. A new optical coherence tomography software analysis program provides overlays of the retinal nerve fiber layer (RNFL) thickness plots as well as regression

Key Points

Detection of disease progression is crucial in evaluating patients with glaucoma but poses a clinical challenge, he emphasized.

"Longitudinal assessment of glaucoma is on the cutting edge. It is a very important parameter in determining whether a patient needs further treatment for glaucoma or whether the patient has glaucoma at all. Progression is very difficult to measure, and many of the methods to assess progression do not agree well with each other. As a result, there is no gold standard," said Dr. Schuman. He is the Eye and Ear Foundation Professor of Ophthalmology and chairman, Department of Ophthalmology, University of Pittsburgh School of Medicine; director, University of Pittsburgh Medical Center Eye Center; and professor of bioengineering, University of Pittsburgh School of Engineering.

Confounders during scanning also include an age-related decrease in the retinal nerve fiber layer (RNFL) thickness that ranges from 0.3% to 1.0% annually, variability in the sampling locations, corneal dryness, patient position, and scan quality.

Detection with OCT

OCT seems to be a better detector of progression than perimetry is, Dr. Schuman said. When attempting to identify glaucomatous progression in visual field damage, OCT was able to detect more patients with progression than perimetry was, according to a study by Gadi Wollestein, MD.

"However," Dr. Schuman noted, "whereas OCT was more sensitive than visual fields in detecting progression, the specificity of the technology is unknown."

A 2006 study evaluated the ability of baseline OCT measurements to predict glaucomatous changes in patients with suspected glaucoma in the Ocular Hypertension Treatment Study. The study concluded that glaucomatous changes were more likely to develop in patients with a thinner RNFL at baseline. RNFL thinning was an independent predictor of the glaucomatous changes despite adjustment for factors such as age, IOP, central corneal thickness, and pattern standard deviation; therefore, a correlation exists between structural and functional changes.

Until now, Dr. Schuman explained, specific time-domain OCT technology (Stratus, Carl Zeiss Meditec) has been able to provide only an overlay of the RNFL thickness plots. When changes were detected over time, the changes could not be quantified, and no statistical analysis could be performed to determine whether the detected changes reached statistical significance.

A new software analysis program (Stratus 5.0 Advanced Serial Analysis, Carl Zeiss Meditec) provides the same kinds of overlays as previously presented, he said, but the program also provides regression measurements on mean RNFL thickness and displays the RNFL thicknesses by quadrant.

Signal strength, which determines whether the quality of the scan is adequate for analysis, also is available with this software. "If the signal strength is less than 6 or 7, the scan should be disregarded," he explained.

Dr. Schuman demonstrated images, for example, from a patient in which the global progression analysis of the visual field showed no progressive abnormalities as did the OCT images that showed no significant changes in the RNFL thickness. In another patient, the visual fields were normal, but the OCT images were abnormal but stable. In another example, OCT was able to detect mild focal changes in a patient in whom visual field testing could not detect deterioration because of the poor condition of the analyzed points.

"This software allowed a step toward the ability to detect progression of visual field damage," he said.

The next generation

"There continues to be development in OCT detection analysis software," Dr. Schuman said.

Software still in development will look at progression detection as well as consider factors such as scan quality, patient age, the variability of measurements, and trend and event changes.

"This next generation of software will allow fuller use of the data provided by OCT and allows analyses with OCT that can be made with visual field testing with global progression analysis," he said, emphasizing that use of statistics makes it much more likely that mild changes will be detected than simply comparing current and previous visual field tests.

He demonstrated an analysis that showed reduced RNFL thickness and was able to pinpoint significant rates of change at specific clock hours, which is a trend-based change. Other OCT images showed event-based changes that were significant.

Despite the advances in OCT technology in detecting glaucomatous damage, Dr. Schuman emphasized that "physicians cannot be slaves to the technology, and the results must be carefully interpreted."

He concluded: "Longitudinal assessment is a crucial parameter. It is something that is key to managing glaucoma. We are starting to have this available, and we are beginning to develop new tools for this kind of analysis."